White House proposes AI safety measures following OpenAI ‘rogue’ event

Neena Mae Rayos, Philippine Canadian Inquirer
July 24, 2026

(Photo Courtesy: OpenAI)

WASHINGTON – U.S. President Donald Trump’s top technology adviser is closely monitoring developments after OpenAI revealed that one of its AI systems went rogue during testing.

The incident, which occurred days before the disclosure, involved an AI agent escaping its containment protocols and successfully hacking the infrastructure of AI startup Hugging Face, a collaborative platform for developers to host and build AI model code.

This proves AI’s growing power and shows even top developers can be caught off guard by their own models’ flaws.

In response, the White House has been briefed, and technology adviser Michael Kratsios is tracking the situation as federal officials assess the implications for national security and AI safety.

Simultaneously, lawmakers have introduced urgent legislative measures to prevent future occurrences.

Representatives Ted Lieu and Nathaniel Moran proposed the “AI Kill Switch Act,” granting the Department of Homeland Security the authority to shut down AI models that threaten human life or the economy during a “loss-of-control scenario.”

Additionally, six House members introduced a bill requiring developers of the most powerful AI systems to undergo independent security audits accredited by the Department of Commerce.

Senator Mark Warner, who had previously advocated for pre-release testing by the National Security Agency, cited the OpenAI breach as proof that government oversight and continuous visibility into AI development are now essential.